In the national initiative, Healthy People 2000, a coordinated effort for the prevention of disease through the improvement of the health of all Americans is underway. An important aspect of the initiative is the role of exercise in the prevention of many diseases, especially osteoporosis and cardiovascular disease (CVD). A recent NIH supported workshop, "Physical Activity and Cardiovascular Health: Special Emphasis on Women and Youth," recommended that additional research is needed, especially in women, concerning the effects of both aerobic and weightlifting exercise on the musculoskeletal and cardiovascular systems. The workshop also emphasized the importance of investigating estrogen replacement therapy (ERT) which has been shown to modify risk factors related to both diseases in postmenopausal women. The investigators propose to conduct a randomized clinical trial employing a 2x2 factorial design to examine the effects of exercise, ERT and their interaction on bone mineral density (BMD) and CVD risk factors in postmenopausal women of age 50 to 65 years. Subjects (n=400) will be randomized to one of four groups: ERT and exercise (n=100); ERT and no exercise (n=100); no ERT and exercise (n=100); no ERT and no exercise (n=100). The exercise program will emphasize aerobic weight-bearing exercises and resistance exercise using free weights and weight resistance machines. The investigators propose to monitor changes in axial and appendicular BMD and total and regional body composition using dual-energy X-ray absorptiometry (DEXA), selected measures of muscle strength, aerobic capacity and indicators of quality of life, along with changes in serum triglycerides, total cholesterol, total HDL and HDL2 and HDL3 subfractions, LDL after 6 and 12 months of training as well as after a one year follow-up in which all subjects will be encouraged to continue the intervention without supervision. Biochemical markers of bone formation and resorption as well as serum levels of calcitropic hormones, sex steroids, sex hormone binding globulin and fasting insulin will also be measured at 6 month intervals during and 12 months after the 1 year intervention to assess hormonal correlates of BMD, serum lipids and lipoproteins. Duplicate blood collections made approximately one week apart will be used to minimize within subject variability in baseline and 12 month lipid and endocrine data. The investigators propose to extend yearly follow-up to 5 years by applying for a continuation. The proposed study uses an integrative approach to examine the effects of ERT, exercise and their interaction on risk factors for two of the major diseases affecting postmenopausal women. Although ERT may reduce the risk for osteoporosis and CVD by maintaining BMD and altering serum lipoprotein profiles, its prescription may not be indicated for all postmenopausal women. Whether exercise is an efficacious alternative or adjunct intervention for risk factor reduction is not presently known. Similarly, it is not known whether the combination of ERT and exercise has synergistic effects on BMD and serum lipids and lipoproteins. The investigators state that the proposed study will increase understanding of the effects of exercise and ERT on bone and lipid metabolism and will provide much needed information for prescriptions of exercise and ERT aimed at the prevention of osteoporosis and CVD in postmenopausal women.